Changes in the Composition and Structure of Glycosaminoglycans in the Human Placenta During Development

Home , Dermatan sulfate, Glycosaminoglycan, Heparan sulfate

Pediat. Res. 7: 965-977 (1973) Chondroitin sulfate placenta glycosaminoglycans

Changes in the Composition and Structure of Glycosaminoglycans in the Human Placenta during Development

TING-YANG LEE, ALEX M. JAMIESON, AND IRWIN A. SCHAFER'44!

Department of Pediatrics at the Cleveland Metropolitan General Hospital, and the Division of Macromolecular Science, Case Western Reserve University, Cleveland, Ohio, USA

Extract Glycosaminoglycans (acid mucopolysaccharides) are ubiquitous in their distribution in the body, yet information as to their biologic function is scanty. Studies of their structure and physical properties in solution suggest that they could function as gel nitration and exchange resins in vivo, thereby playing an important role in regulation of the passage of molecules through the ground substance of connective tissue. The glycosaminoglycan(s) (GAG) composition of the human placenta and the molecular structure of specific GAG has been studied by chemical, enzymatic, and physical methods at 12-18 weeks and at 40 weeks gestational age to explore this postulated structure-function relation. The young placenta contained more GAG (222 mg/100 mg dry defatted tissue) than did the term placenta (155 mg/100 g dry defatted tissue). Sulfated GAG comprised 56% of the total GAG in the young placenta versus 74% in the term placenta due to increased concentrations of dermatan sulfate (25% term versus 13% young placenta) and heparan sulfate (22% term versus 15% young placenta). Ghondroitin was a major component in the young placenta and comprised 22% of the total GAG, whereas the term placenta contained only 9%. Both young and term placenta showed similar quantities of hyaluronic acid and chondroitin 4- and 6-sulfates. Ghondroitin sulfate from the young placenta differed from the polymer in the term placenta in that it contained a higher proportion oi unsulfated dissaccharides. Differences were also found in the molecular structure of dermatan sulfate. Hyaluronidase digestion of purified dermatan sulfate from young placenta produced a 50% reduction in average molecular weight compared with a 30% reduction in the molecular weight of dermatan sulfate isolated from term placenta. The smaller molecular weight fragments of dermatan sulfate from young placenta indicate differences in molecular structure due either to the number of glucuronic acid substitutions or to their position in the polymer chain, or to changes in the concentration of hybrid molecules.

Speculation The age-related changes in the composition and molecular structure of placental glycosaminoglycans will result in ground substance gels of differing physical properties. This could alter the transport of molecules through placental connective tissue and affect the rate of fetal growth.

965 966 LEE, JAMIESON, AND SCHAFER

Introduction mal human placenta at 12-18 weeks gestational age as Ground substance is a histochemical descriptive term compared with the term placenta at 40 weeks, and to referring to the material of connective tissue which is examine the molecular structure of specific GAG. Our neither cell nor fiber. It is the matrix in which cells data indicate that the concentrations of several GAG and fibers are embedded and through which metabo- change dramatically as this organ matures, and that lites pass in exchanges between cells or between cells the molecular structure of dermatan sulfate and chon- and the vascular system. Chemically, GAG (acid muco- droitin sulfate is also altered. polysaccharides) comprise a major and characteristic molecular species of this matrix [31]. Glycosamino- Materials glycans may be described as protein-linked linear flexi- Two pools of term placentas from 40-week pregnancies ble polymers, composed of disaccharide repeating units were obtained from the Perinatal Research Unit and containing 1 or more anionic groups per repeat unit. the Department of Obstetrics of the Cleveland Metro- They occur in different tissues in typical patterns politan General Hospital. Initial analyses were carried which may change with maturation and ageing. Infor- out on three whole placentas designated as pool A. mation on their biologic function is scanty [26]. These data were further confirmed by analysis of 20% The structure of GAG as related to their physical wet weight samples from 10 placentas designated as properties has been extensively studied in vitro. Based pool B. The clinical course in all 13 women was nor- on these types of experiments, several investigators mal throughout gestation. Their infants were normal have proposed that GAG play an important physio- at birth and weighed between 2,700 and 3,850 g. The logic role in the regulation of transport exchange wet weight of the placentas ranged between 450 and through the ground substance by virtue of their charge 800 g. [32] and steric configuration [20]. Twenty-two placentas between 12 and 18 weeks of All glycosaminoglycans are anionically charged at gestational age were obtained through legalized thera- physiologic pH and act like polyelectrolytes; therefore, peutic abortions by abdominal hysterotomy at the they may function as exchange resins in vivo, and bind University of Helsinki Women's Central Hospital. compounds or ions which are cationically charged [25]. These specimens weighed between 4 and 22 g. The As polymers in solution, it has been shown that pregnancies had been interrupted for psychiatric rea- their steric configuration may exclude or sieve mac- sons, none of the mothers had organic disease, and romolecules traveling through the domain which they their fetuses appeared to be normal. The placentas occupy. This property in vivo could influence the dis- were frozen, stored, and transported in Dry Ice. They tribution of extracellular protein between various tis- were stored in our laboratory at —60° until analysis. sues, their colloid osmotic pressures, and the transport of macromolecules through and between tissue com- Methods partments [19]. Whole young and pool A term placentas were proc- There is no direct experimental evidence to estab- essed in parallel for all succeeding steps. Pool B term lish this postulated structure-function relation in vivo. placentas were analyzed separately using the same pro- One way to explore this relation would be to establish cedures except that only 20% of the weight of each a coiTelation between the composition and structure of placenta was used for the tissue pool. GAG in a tissue as related to the transport function of The umbilical cord, fetal membranes, and the large that tissue. The placenta seemed a good organ to study vessels on the chorionic surface were removed by experimentally for the following reasons: (2) it regu- dissection. Tissues which remained were minced in a lates the transport of metabolites between distinct ana- meat grinder and pooled. The total wet weight of the tomic and physiologic compartments [1], (2) its trans- tissue pool from the young placenta was 152 g, from port functions change with maturation and in certain the term placenta pool A, 1,154 g, and term placenta pathologic states [21], and (3) if changes in chemical pool B, 741 g. composition or the structure of component GAG are The minced tissue pools were washed repeatedly related to transjDort function, this may be reflected in with 75% ethanol and the wash removed by centrifu- the size of the infant or in his physiologic status at gation at 2,000 rpm for 20 min. This process was re- birth [21]. peated until the wash became colorless. The tissues The purpose of this study is to establish base-line were then defatted with acetone-ether (1/1, v/v) at 4° data on the glycosaminoglycan composition of the nor- for 16-18 hr and the insoluble material was collected Developmental changes in placental glycosaminoglycans 967 by centrifugation at 2,000 rpm for 20 min at 4°. This Table I. Yield for glycosaminoglycan (GAG) after second process was repeated 4-5 times in young placenta and Pronase digestion

10 times in the term placenta before cloudiness disap- Defatted Crude GAG Crude GAG tissue, powder, dry powder, uronic peared completely from the supernatant. dry wt, g wt, mg acid content, The tissue pools were then washed with 95% mg ethanol followed by diethyl ether and dried under a Term placenta vacuum to a constant weight. The yields were 147 g of Pool A 60 518 90 dry defatted tissue from the 3 term placentas, pool A; Pool B 15 240 24 102 g from the term placenta, pool B; and 17.2 g from Young placenta 10.5 128 23.3 the 22 young placentas. The dry defatted tissues were powdered in a Waring Blendor. columns (1.5 by 30 cm) of Dowex 1-X2, chloride form, 200-400 mesh resins. The columns were washed with Isolation of Glycosaminoglycan 200 ml distilled water. Separation of the GAG mixture Next 60 g dry defatted term placenta pool A pow- was achieved with a stepwise gradient of sodium chlo- der, 15 g term placenta pool B powder and 10.5 g ride from 0.25 through 4.0 M. Twelve-milliliter aliquots young placenta powder were suspended in 0.5 M sodium of eluate were collected and the uronic acid content acetate and digested with Pronase B [35] at 65°, and was measured. Only after elution was complete at a GAG were recovered from the digest by the procedure given chloride concentration was the concentration in- of Svejcar and Robertson [33] with only two minor creased by 0.25 M increments and the elution repeated. modifications. Trichloroacetic acid was added to a con- Tubes which corresponded to each peak were pooled, centration of 10% and GAG and glycopeptides were dialyzed in viscose cellulose bags against distilled water recovered from 80% ethanol. The precipitates were to remove salt, and evaporated to dryness. This mate- washed with 95% ethanol and diethyl ether and dried rial was solubilized in 10% sodium acetate and the under a vacuum. GAG recovered from 80% ethanol at 4° after 16 hr, The dried precipitates were suspended in 0.01 M washed with 95% ethanol and then diethyl ether, and phosphate buffer in 0.02 M sodium chloride at pH 7 dried under vacuum to a constant weight. and digested with hog pancreas diastase [36], 6.5 mg/g The recovery of uronic acid applied to the columns tissue at 37° for 24 hr. Trichloroacetic acid, 100%, was was 75% in the term placenta pool A, 92% in term added to the digest to a final concentration of 10%, placenta pool B, and 86% in the early placenta. After the precipitate was removed by centrifugation at 2,000 dialysis and precipitation, the recoveries were 53% rpm for 20 min at 4°, and GAG were recovered from term placenta pool A, 84% term placenta pool B, and the supernatant with 80% ethanol which contained 59% in the young placenta. The lower total recoveries 5% potassium acetate. This precipitate was washed in the term placenta pool A and the young placenta successively with 95% ethanol and diethyl ether and were not due to losses during dialysis but occurred dried under a vacuum. because of the size and shape of the tubes that were To remove any remaining proteins, a second Pro- used for precipitation and drying of these samples. We nase digestion was carried out and the GAG recovered were unable to recover the dry powder completely as described above. The final precipitate was solubi- from these vessels. In term placenta pool B, recoveries lized in 10% sodium acetate, then reprecipitated with were higher because the samples were weighed and 80% ethanol for 16 hr at 4°, and collected by centrifu- resolubilized in the same tube. As indicated by our gation at 2,000 rpm for 20 min. The precipitate was analytic results, the GAG composition of the dry pow- washed sequentially with 95% ethanol and diethyl der recovered from pool A was similar to the GAG ether and dried under a vacuum to constant weight. composition of pool B. Therefore, we do not believe The yields were as shown in Table I. that differential losses of GAG occurred during isolation procedures. Purification of Glycosaminoglycan Column chromatography. Term placental pool A Isolation of Pure Dermalan Sulfate (195 mg), term placental pool B (228 mg), and young To carry out ethanol fractionations, additional placental GAG mixtures (98 mg) (which contained 34 crude GAG powder was prepared as described above mg, 23 mg, and 18 mg uronic acid, respectively), were from 6.3 g moisture-free defatted young placenta (yield each dissolved in 5 ml distilled water and applied to 104 mg). This was combined with 22 mg material 968 LEE, JAMIESON, AND SCHAFER which remained from the first preparation. The 123 covered by centrifugation at 2,000 rpm for 20 min. mg so obtained was solubilized in 5% calcium acetate Both the precipitate, and the supernatant which was in 0.5 M acetic acid, as was 166 mg of GAG powder evaporated to dryness under reduced pressure, were from the term placenta. The mixture of GAG was solubilized in 0.001 M sodium hydroxide and analyzed separated as calcium salts with concentrations of for hexosamine and uronic acid. Standard GAG were ethanol at 20%, 40%, and 60% as described by Meyer used each time for control [38]. et al. [27]. Each fraction was treated by the alkaline Chondroitinase ABC [39] digestions were carried out copper precipitation method of Cifonelli et al. [6]. following method I of Saito [30]. The completeness of Dermatan sulfate was obtained as its copper complex digestion was checked with GAG standards after 1- from the 20% ethanol fraction and converted to its and 5-hr incubations at 37° using 0.2 U enzyme/100 [xg sodium salt. This precipitate weighed 19.7 mg from GAG initially, with the addition of 0.1 U enzyme after the term placenta and 8.3 mg from the young placenta. 2.5 hr. Five hours were required to completely digest The analytic data from the purified dermatan sul- hyaluronic acid whereas the other GAG were com- fate fractions are shown under Results in Table V. pletely digested after 1 hr. The total amounts of GAG in each fraction that were resistant or susceptible to Analytic Methods digestion with chondroitinase ABC following 5-hr in- Chemical methods. Uronic acid was determined by cubations were determined by recovering the undi- the borate-carbazole method of Bitter and Muir [2], gested material as a precipitate from 80% ethanol which and also by the naphthoresorcinol method of Pelzer contained 1% potassium acetate after 16 hr at 4°. The and Staib [28] using glucuronic acid as the standard; supernatant was evaporated to dryness under a vac- hexose by the anthrone method using D-galactose as uum. Both the precipitate and residue from the super- the standard; hexosamine by the method of Boas [3] natant were solubilized in 0.001 M sodium hydroxide omitting the Dowex 50W-X8 step; and the differentia- and their uronic acid content was measured. The dition of giucosamine and galactosamine by the method gests were then chromatographed on Whatman no. 1 of Ludowieg and Benmaman [24]. The samples for (11.5 by 18.5 inch) paper from application spots 0.5 cm hexosamine determination were hydrolyzed in 2 N hy- in diameter. Samples were first desalted by descending drochloric acid at 100° for 10 hr, and both D-glucosa- chromatography with 1-butanol-ethanol-water (52/32/ mine-HCl and D-galactosamine-HCl were used as stan- 16 v/v) for 24 hr, and the unsaturated disaccharides dards. Sulfate was determined by the barium chloride were then separated by descending chromatography in turbidimetric method of Dodgson and Price [7], and 1-butyric acid-0.5 M ammonia (5/3 v/v) for 48 hr at 2-deoxy-2-sulfaminohexose was determined by the room temperature. The spots were identified under method of Lagunoff and Warren [18] using D-giucosa- ultraviolet light by comparison with standard unsatu- mine-HCl as standard. Protein was measured by a rated disaccharides [40], cut from the paper, eluted in modification of the method of Lowry et al. [23] in 2.0 ml of 0.01 M HC1 at 50° for 10 min, and their 2 which the CuSO4 was stabilized in alkaline solution ultraviolet absorption was measured in a DU Beck- with disodium EDTA [29]. man spectrophotometer [41] at 232 nm. The amounts Eleclrophoresis. Electrophoresis was performed on of each unsaturated disaccharide recovered from the cellulose acetate membrane strips in 0.2 M zinc sulfate paper were calculated using the following millimolar as described by Haruki and Kirk [12]. The strips were absorption coefficients: 5.7 for nonsulfated disaccha- stained with 0.1% Alcian blue in 5% acetic acid-20% rides, 5.1 for 4-sulfated disaccharides, and 5.5 for 6-sul- ethanol. fated disaccharides. Enzymatic methods. Hyaluronidase digestions were Chondroitinase AC II digestions were carried on se- carried out at 37° for 24 hr in NaCl-acetate buffer, pH lected chondroitin sulfate fractions to differentiate 5.4, using 150 USP U enzyme/100 p,g GAG as de- chondroitin sulfate from dermatan sulfate using the scribed by Saito [30]. Trichloroacetic acid, 100%, was methods outlined above [39]. The percentage of GAG added to the digest to a final concentration of 10%, digested with chondroitinase AC II was calculated and the precipitate was removed by centrifugation. from the ultraviolet absorption of disaccharides re- Four volumes of absolute ethanol containing 1% po- covered after paper chromatography. tassium acetate were added to the supernatant, and the Measurement of molecular weight. The molecular precipitate which formed at 4° after 16 hr was re- weight of dermatan sulfate was measured by optical Developmental changes in placental glycosaminoglycans 969 mixing spectroscopy, a technique developed for analy- shape computer analysis is nearing completion and sis on quasielastic light scattering, in which the spec- will be published. troscopic distribution of Rayleigh-scattered light is measured with allowance for calculation of diffusion Results coefficients and hence molecular weights. Two-milligram quantities of purified dermatan sulfate were dis- Column Chromalography solved in 20 ml distilled deionized water. The hetero- Chemical and enzymatic analysis of GAG fractions dyne Rayleigh spectrum of these solutions was recorded obtained by column chromatography from the young in a Brice Phoenix light-scattering cell no. C-105 and term placenta are presented in Tables II and III. [42] at room temperature (27°) using a laser heter- The yield of GAG, as percentage of distribution of odyne spectrometer which has been described in detail uronic acid recovered from the columns, differed in [14]. Weight average molecular weights were obtained the young and term placenta in the 0.5 M and 1.5 M by interpolation from known standards. A preliminary fractions (Table II). The 0.5 M fraction was larger in report of the procedure for the measurement of molec- the young placenta as compared with the term pla- ular weights of highly sulfated GAG has appeared [13]. centa, whereas the reverse was true in the 1.5 M frac- A more detailed molecular weight study using line tions. The mixture of GAG applied to the column was

Table II. Analytic data for young and term placental glycosaminoglycan fractions from Dowex 1 column1

After Hyaluronidase Digestion Fractions Uronic Sulfaminc- Glucosamine to Borate to Naph- Borate to Naphtho- M NaCl Yield'2) Acid Hexose Sulfate hexcse Galactosamine thoresorcinol Reffaim'rcw) resorcinol Molarity (moles/mole of total hexosamine) ratio ratio (::) ratio

Young Placenta

0.25 4.91 0 .15 1 39 0 0. 57:43 1.74 100

0.5 26.99 0 .99 0 46 0 0. 64:36 6.04 16 0.51

0.75 10.25 1 39 0 40 0 53 0 20 67:33 5.78 61 3.83

1.0 9.49 1.29 0 33 0 77 0 18 86:14 8.05 66 3.16

1.25 10.10 1.50 0 24 1 28 0 P5 25:75 6.20 35 2.65

1.5 28.11 1.12 0 32 1 47 0 7:93 2.21 42 0.78

1.75 7.01 1 .16 0 37 1 37 0 10:90 1.49 51 0.5S

2.0 1.99 1.28 0 27 1 67 0 13:87 1.29 66 0.54

3.0 0.32 1.36 1 22 2 80 0 35:65 1.29 -

4.0 0.78 0 .71 0 65 1 19 0 94:6 3.03 - -

Term PIacenta A B A B A B A B A B A B A B A B A B

0.25 3.93 3. 42 0.16 0.12 1 54 1.00 0.06 0.03 0. 0. 79:21 86:14 0.98 3.86 100 100 -

0.5 13.61 12. 53 0.97 0.80 0 69 0.73 0.15 0.01 0. 0. 74:26 57:43 3.70 2.87 31 22 0.83 0.25

0.75 11.16 12. 58 1.41 1.65 0 79 0.54 0.64 0.63 0.10 O.lf 77:23 64:36 4.34 4.86 75 60 4.36 4.04

1.0 13.87 14. 04 1.39 1.59 0 62 0.32 1.10 0.83 0.34 0.3? 84:16 83:17 5.61 7.09 79 76 4.04 5.22

1.25 10.74 8. 48 1.73 1.57 0 50 0.33 1.23 1.32 0.11 0.13 49:51 13:87 6.64 4.82 51 40 2.43 3.09

1.5 37.09 35. 89 1.15 1.27 0 34 0.20 1.44 1.36 0. 0. 14:86 0:100 1.98 1.95 57 48 0.88 0.95

1.75 6.64 9. 29 1.04 1.26 0 40 0.20 1.44 1.48 0. 0. 10:90 0:100 1.11 1.37 76 69 0.83 0.84

2.0 1 .80 1. 86 0.81 0.93 0 61 0.29 1.37 1.37 0. 0. 26:74 12:88 0.89 1.01 85 80 0.64 0.64

3.0 0.38 1. 72 0.68 0.98 1 18 0.53 1.72 1.31 0. 0. 59:41 20:80 1.01 1.35 - 79 0.63

4.0 0.90 0. 18 0.90 - 0 72 - 0.51 - 0. - 59:41 1.77 - - - - 1 A: sample pool of 3 whole placentas; B : sample of 10 placentas using 0.20 wet weight of each for pool. 2 Percentage of distribution for total glycosaminoglycan recovered from columns after dialysis and precipitation with 80% ethanol based on borate carbazole uronic acid content. borate uronic acid content after digestion 3 Percentage remaining : X 100. total borate uronic acid content 970 LEE, JAMIESON, AND SCHAFER

Table III. Ghondroitinase ABC and AC-II digestion of young and term placental glycosaminoglycan fractions from Dowex 1 column1

Dis tribution of Digested GAG 5 of total uronic acid in each fraction (4)

% of total GAG Digested<3) Fractions Yield unsaturated disaccharides Hyaluronic Acid NaCl Chondroitin - Chondroitin- Molarity % (2) ase AC-II ase ABC non-sulf ated 4-sulfe ted 6-sulfated

Young Placenta

0.5 26.99 81 51 0 0 30

0.75 10.25 - 42 33 2 trace 7

1.0 9.49 - 39 18 trace 10 11

5 1.25 10.10 70 70 15(16) 9(8) 46(46) 0(0)

1.5 28.11 69 100 5(5) 37(10) 58(54) 0(0

1.75 7.01 53 100 trace (8) 55(5) 45(40) 0(0

Term Placenta A B B A B A B A B : A B A B

0.5 13.61 12.53 - 67 75 34 41 0 0 0 0 33 34

0.75 11.16 12.58 - 28 39 19 19 0 0 0 0 9 20

1.0 13.87 14.04 20 24 7 6 trace trace 1 3 12 15

1.25 10.74 8.48 62 61 61 11 8(8) 13 18(17) 26 35(37) 11 0(0)

1.5 37.09 35.89 58 100 100 0 1(2) 54 54(15) 46 45(41) 0 0(0)

1.75 6.64 9.29 25 100 100 0 0(0) 72 73(7) 28 27(18) 0 0(0)

2.0 1.80 1.86 - 78 100 0 0 58 75 20 25 0 0

3.0 0.83 1.72 - - 76 - o ~ 52 - 24 - 0 1 A: sample pool of 3 whole placentas; B : sample pool of 10 placentas using 0.20 wet weight of each for pool. 2 Percentage distribution of total glycosaminoglycan recovered from Dowex 1 columns after dialysis and precipitation with 80% ethanol based on borate carbazole uronic acid content. 3 Each fraction was digested with chondroitin ABC for 5 hr, and the undigested material was precipitated with 80% ethanol at 4° for 16 hr. Borate uronic acid was measured in the precipitate and supernatant. Chondroitin AC II digestions were carried out for 5 hr and the percentage of digested glycosaminoglycan calculated from the ultraviolet absorption of disaccharides recovered after their separation by chromatography. 4 The concentration of unsaturated disaccharides was calculated from their ultraviolet absorption after their separation by paper chromatography. Hyaluronic acid was calculated as the difference of the borate uronic acid content of the total chondroitinase ABC- digested glycosaminoglycan and the recovered unsaturated disaccharide. 6 Results of chondroitinase AC II digestion are shown within parentheses. resolved on the basis of anionic charge which is best Characterization and Quantitation of Component Gly- correlated with the total sulfate content (Table II). cosaminoglycan in Dowex 1 Fractions The term placenta had a higher proportion of sulfated In order to quantitate the concentration of specific GAG than did the young placenta, comprising 71 % of GAG in each fraction, enzyme digestions were carried the total GAG in the term versus 58% in the young, as out using bovine testicular hyaluronidase and chon- judged by GAG eluted at 1.0 M NaCl or above. droitinase AC II and ABC. Chemical analyses and Based on the ratio of glucosamine to galactosamine, electrophoresis were done on the GAG which were total sulfate, sulfaminohexose, and hexose content (Table II), each fraction from the column contained a enzyme resistant or digestible, and the proportion (per- mixture of two or more GAG. This was confirmed by centage) of a specific GAG in each fraction was calcu- cellulose acetate electrophoresis. Fractions 0.25-0.5 M lated based on these data. showed glycopeptide, hyaluronic acid, and undersul- Fraction 0.25 M. One hundred percent of the mate- fated chondroitin. Fractions 0.75-1.25 M contained hep- rial was hyaluronidase resistant. The precipitate con- aran sulfate and chondroitin sulfate isomers while tained a high hexose content, little uronic acid, no fractions 1.5 through 2.0 M contained dermatan sulfate sulfate, and was designated as glycopeptide because and chondroitin sulfate isomers. The 3.0 and 4.0 M 13% of its dry weight was composed of protein. fractions contained increasing amounts of hexose, glu- Fraction 0.5 M. The very low sulfate content in this cosamine, and sulfate, which suggested the presence of fraction indicated that it contained mainly nonsul- a keratan sulfate-like material. fated GAG. Hyaluronidase and chondroitinase ABC Developmental changes in placental glycosaminoglycans 971

Term Placenta - 4O Weeks Dow§xI Standard Fraction Disocchorides .5 .75 1.0 1.25 1.5 1.75 2.0 6-S * I 9 » • 4-S i •

0-S

HyolLjronic Acid I

Young Placenta -12-18 Weeks 6-S V 'lift 4-S I « t 0-S kit-

Hyaluronic Acid

Fig. 1. Chromatography of unsaturated disaccharides after chondroitinase ABC digestion (tracings of the original chromatograms). digestions yielded similar amounts of undigested mate- gestible materials and nonsulfated chondroitin was rial (Tables II and III). The enzyme-resistant fraction considered to be hyaluronic acid (Table III) and was was considered to be giycopeptide despite the low car- 33% and 34% in the term placenta versus 30% in the bazole-naphthoresorcinol ratio because digestion with young placenta. chondroitinase ABC eliminated dermatan sulfate from Fractions 0.75—1.0 M. The high glucosamine and sul- the precipitate. In addition, only a single band corres- faminohexose content of these fractions indicated that ponding to standard hyaluronic acid was found on they contained heparan sulfate. The total sulfate to cellulose acetate electrophoresis. Paper chromatogra- hexosamine ratio was low, which suggested either an phy of the chondroitinase ABC digest demonstrated undersulfated heparan sulfate or the presence of other two spots (Fig. 1), which corresponded to the standard under- or nonsulfated GAG (Table II). After digestion nonsulfated disaccharide, and to that obtained with with hyaluronidase, the borate to naphthoresorcinol standard hyaluronic acid degraded with the same en- ratio remained high, which eliminated dermatan sul- zyme. Nonsulfated chondroitin was calculated on the fate as the resistant GAG in the precipitate (Table II). basis of the nonsulfated disaccharide recovered by the This was confirmed by chondroitinase ABC digestion paper chromatography and comprised 51% of the total which proved clearly that the enzyme-resistant GAG uronic acid of this fraction in the young placenta ver- was heparan sulfate which comprised 76 and 69% of sus 34% and 41% in the term placenta (Table III). the total uronic acid in the term placenta and 59% in The difference between total chondroitinase ABC-di- the young placenta (Table III). 972 LEE, JAMIESON, AND SCHAFER

Term Placenta -40 Weeks Youna Placenta -•12-18 Weeks Dowex 1 Standard Fraction Disccchoride i.25 1.5 1.75 1.25 1.5 1.75

6- S N N M 1 1 » i • . i 4-S * o-s 1 1 \ 1

Fig. 2. Chromatography of unsaturated disaccharides after chondroitinase AC II digestion (tracings of the original chromatograms).

The major disaccharide isolated and recovered after sulfate; heparan sulfate and nonsulfated chondroitin chromatography of the chondroitinase ABC digest was contributed smaller amounts (Table III). a nonsulfated disaccharide in both the term and young- Fractions 1.5-1.75 M. Galactosamine was the princi- placenta (Table III), with smaller quantities of 6-sul- pal hexosamine in these fractions (Table II) and their fated disaccharide, and only trace amounts of 4-sul- GAG were completely digested to unsaturated disac- fated material (Fig. 1). charides by chondroitinase ABC, which indicated that Fraction 1.25 M. The increase of galactosamine indi- they were isomers of chondroitin sulfate (Table III). cated that the major GAG in this fraction were the Of the GAG, 42-76% were resistant to hyaluronidase isomers of chondroitin sulfate. The presence of sulfa- and chondroitinase AC II (Table II and III). These minohexose indicates the presence of heparan sulfate isomers showed a low borate to naphthoresorcinol (Table II). There were no differences found in the ratio with no sulfaminohexose detected, which indi- susceptibility of the GAG in this fraction to digestion cates that the hyaluronidase-resistant isomers were der- with hyaluronidase and chondroitinase AC II and matan sulfate (Table III). The major disaccharides ABC, which indicates that the enzyme-resistant GAG recovered after digestion with chondroitinase ABC was heparan sulfate (Tables II and III). were 4- and 6-sulfated compounds, whereas after chon- Three unsaturated disaccharides were obtained by droitinase AC II digestions, 6-sulfated disaccharides the paper chromatography after chondroitinase AC II predominated. Hyaluronidase- and chondroitinase AC and ABC digestion (Figs. 1 and 2). The susceptibility II-resistant GAG were designated as dermatan sulfate. to digestion with chondroitinase AC II shows clearly The 4-sulfated and 6-sulfated disaccharides recovered that all of the 4-sulfated disaccharides were derived after chondroitinase AC II digestion were derived from chondroitin 4-sulfate (Table III). from chondroitin 4-sulfate and 6-sulfate. Small quanti- The principal GAG in this fraction was still hep- ties of nonsulfated disaccharides were present in the aran sulfate in the term placenta with smaller 1.75 M fraction in the young placenta whereas the term amounts of chondroitin 6-sulfate, nonsulfated chon- placenta contained none (Fig. 2). A light spot of very droitin, and hyaluronic acid. By contrast, the major slow RP was present in each fraction in both the term component in the young placenta was chondroitin 6- and young placenta. These were considered to be di- Developmental changes in placental glycosaminoglycans 973

sulfated disaccharides but their concentration was so compounds. Insufficient sample was available for fur- small that they were not considered in this composi- ther characterization of these fractions. tional analysis. Table IV summarizes the composition of GAG in In the term placenta, dermatan sulfate composed the term and early placenta. The amounts of each 57% and 49% of these fractions and chondroitin 6-sul- GAG identified in the column fractions have been fate 43% and 36%, whereas in the young placenta summed to give this composite picture. chondroitin 6-sulfate comprised 52% of the total with The GAG composition is presented as percentage of smaller amounts of dermatan sulfate (35%). Only distribution of total GAG recovered from the columns small amounts of chondroitin 4-sulfate were found in after dialysis and reprecipitation of the GAG. Definite either term and young placenta. differences were found. The early placenta had a Fraction 2.0 M. In both the term and young placenta higher proportion of nonsulfated chondroitin, whereas this fraction showed material resistant to digestion the proportions of dermatan and heparan sulfate were with hyaluronidase with a low borate-naphthoresor- increased in the term placenta. Chondroitin 6-sulfated cinol ratio which indicates the presence of dermatan was the principal hyaluronidase-susceptible chondro- sulfate (Table II). Sufficient material for chondroitin- itin sulfate isomer in both the young and term pla- ase digestion was not available in the young placenta. centa, with only small quantities of chondroitin 4-sul- In the term placenta, 78% of the material was digesti- fate detected. Overall, the young placenta contained ble with chondroitinase ABC and the major disaccha- more GAG than did the term placenta, but the term ride was 4-sulfated. Approximately 66% of this frac- placenta had appreciably higher proportions of hyalu- tion was attributed to dermatan sulfate in the young- ronidase-resistant sulfated GAG. There appeared to be placenta on the basis of resistance to hyaluronidase a reciprocal relation between the proportion of non- and borate-naphthoresorcinol ratio (Table II). In the sulfated chondroitin and the concentrations of derma- term placenta dermatan sulfate comprised 58 and 75% tan and heparan sulfate. on the basis of chondroitinase ABC digestion. (Table Structural analyses of dermatan sulfate from young III, Fig. 1) In addition, in the term placenta, the con- and term placenta. The chemical composition and en- centration of glucosamine and hexose increased, with zymatic analysis of dermatan sulfate isolated from the decreased content of uronic acid suggesting that a ker- term and young placenta is given in Table V, and atan sulfate-like material was eluted with this fraction. compared with standard dermatan sulfate isolated No further characterization of this compound was car- from hog mucosa. These data indicate that the GAG ried out because of the limited amounts available for isolated from the placenta was highly purified derma- analysis. tan sulfate. After digestion with testicular hyaluroni- Fractions 3.0-4.0 M. AS the molarity increased, the dase no differences were detected by chemical and en- eluates contained increasing quantities of glucosamine zymatic analysis. Molecular size as measured by optical and hexose with smaller molar ratio for uronic acid, mixing spectroscopy showed a striking reduction in the which suggested the presence of keratan sulfate-like molecular weight of the dermatan sulfate isolated from

Table IV. Glycosaminoglycan composition in normal term and young placenta

Recovery of uronic acid from Dowex 1 Distribution, %l Uronic acid column in dry defatted tissue, mg/100 After In dialysis Hyal- Chon- eluate, and pre- Glyco- uronic Dermatan Heparan Keratan % cipitation, peptide acid chondroitin 4-sulfate sulfate sulfate sulfate %

Term placenta2 Pool A 150 75% 53% 8.42 8.33 8.90 22.21 1.40 26.37 23.33 Some Pool B 160 92% 84% 6.55 8.89 9.77 20.83 7.47 24.33 21 .56 Some Young placenta 222 86% 59% 10.04 9.86 22.45 24.26 4.18 13.58 14.77 Some 1 Each Dowex 1 chromatographic fraction was studied chemically before and after digestion with hyaluronidase, as well as chondroitinase ABC and AC II. The identity and amount of specific GAG were calculated from these data. 2 pool A: sample pool of 3 whole placentas;pool B: sample of 10 placentas using 0.20 wet weight of each for pool. 974 LEE, JAMIESON, AND SCHAFER

Table V. Analyses of placental dermatan sulfate (DS) before (A) and after (B) treatment with testicular hyaluronidase1

Unsaturated disaccharide Borate Weight, % after chondroitinase -tiyaiu- to ABC digestion eight Yield, ronidase naphthore- Molecular \v average rfc SE resistant, sorcinol i 4- ratio Hexos- [Ironic Sulfate Disul- 6-sul- sul- amine acid fated fated fated

Hog mucosal DS A 0.77 28 26 (0.87) 23 (1 .51) 7.8 4.7 87.5 27,900 ± 1,800 B 97 0.69 33 32 (0.87) 24 (1 .33) 8.8 2.4 88.8 24,500 ± 1,800 Term placental DS A 14.82 0.76 33 31 (0.89) 24 (1 .37) 0.9 6.4 92.7 23,400 ± 1,500 B 94 0.75 26 27 (0.98) 19 (1 .37) 0.5 0.9 98.6 15,700 ± 1,500 Young placental DS A 8.75 0.73 22 27 (1.12) 18 (1 .48) 1.5 6.7 91.8 21,700 ± 1,500 B 97 0.70 27 27 (0.91) 19 (1 .34) 0.5 0.5 99.0 10,000 ± 1,500 Figures in parentheses are data expressed on molar ratio basis to hexosamine. Percentage of total uronic acid in the crude placental GAG based on borate carbazole uronic acid. the placenta, whereas hog mucosal dermatan sulfate the term or young placenta, reported no age-related was resistant to the action of this enzyme. The average change in the concentrations of dermatan sulfate, but size of the polymer was decreased by 50% in the young did show that the young placenta contained more un- placenta and by 30% in the term placenta. dersulfated chondroitin 6-sulfate. These studies were carried out on the central area of the placenta midway Discussion between the decidua and chorionic plate while we used the whole tissue. The difference in results may be These data show that age-related changes occur in the attributable to differences in methodology but not to composition and structure of GAG in the human pla- tissue sampling since Takeuchi and Schafer [34] centa. showed that the midportion of the placenta contained The young placenta (12-18 weeks) and term pla- heparan sulfate. centa (40 weeks) both contain hyaluronic acid, chon- Our analysis identified two types of chondroitin pol- droitin 6-sulfate, chondroitin 4-sulfate, nonsulfated ymers in the placenta. Over 76% of the nonsulfated chondroitin, dermatan sulfate, and heparan sulfate, disaccharides were recovered from the 0.5 M and 0.75 but in differing proportions. Sulfated GAG comprised M column fractions which contained little or no sulfate. 56% of total GAG in the young placenta as compared The young placenta contained more of these unsulfated with 74% in the term placenta with differences ac- chondroitin chains than did the term placenta. Chon- counted for by an increase in hyaluronidase-resistant droitin 6- and 4-sulfate isolated from the higher salt GAG and dermatan and heparan sulfate. Only small column fractions were found to be hybrid molecules amounts of chondroitin 4-sulfate were detected. In ad- composed of sulfated and nonsulfated disaccharides. dition, analytic data suggest the presence of small The young placenta showed a larger proportion of quantities of keratan sulfate-like GAG. chondroitin sulfate polymers with this structure than Calatroni and Di Ferrante [5] reported similar re- did the term placenta. sults for the GAG composition of term placenta except Kaplan and Meyer [16] reported that human aorta for larger amounts of chondroitin 4-sulfate. This dis- contained hyaluronic acid, heparan sulfate, dermatan crepancy may be due to the methods used to identify sulfate, and chondroitin 6-sulfate but no chondroitin chondroitin 4-sulfate. They used infrared spectral 4-sulfate. This striking similarity in the overall GAG analysis which does not differentiate mixtures which composition of both tissues is also paralleled by time contain 4-sulfated isomers of chondroitin from derma- dependent changes in their composition. Both tissues tan sulfate while the enzymatic methods used in this show decreasing total GAG content as they age with study clearly separate mixtures of these two isomers. proportional increases in sulfated GAG. The increase The only published study which compares the GAG in sulfated GAG in the placenta is due to increased composition of young and term placenta was by Lovell concentrations of dermatan sulfate and heparan sul- el al. [22] who did not detect heparan sulfate in either fate, with reciprocal decreases in the concentrations of Developmental changes in placental glycosaminoglycans 975 nonsulfated chondroitin. The aorta shows similar age- nique of molecular weight measurements using optical related increases of heparan and dermatan sulfate con- mixing spectroscopy, when coupled with chemical or centrations with decreased concentrations of chondro- enzymatic cleavage methods, provides a powerful tool itin 6-sulfate and hyaluronic acid. Because the pla- with which to study structural differences in GAG and centa is a highly vascular organ, the similarity of GAG possibly other types of polymers and proteins in em- composition to the aorta may simply reflect this fact. bryonic tissues. The main virtues of the method are Although the patterns of specific GAG are different, that only milligram quantities of sample are needed striking age-dependent changes have also been re- for analysis and the measurements do not destroy the ported for human costal cartilage [15] and skin [4], sample, which can then be utilized for other analytic which suggests that GAG metabolism is related to bio- procedures. logic aging in many tissues. The biologic implications of the compositional and Age-dependent changes in the structure of a specific structural changes in GAG found in the placenta at GAG have not been previously described. Dermatan 12-18 weeks and 40 weeks of gestational age remain sulfate, the GAG selected for structural analysis in this speculative. Theoretically, the increased proportion of study, is composed of /?(l-^4)-linked L-iduronic «-l,3- sulfated GAG in the 40-week placenta might be ex- Ar-acetyl-D-galactosamine disaccharides. Dermatan sul- pected to change both the physical and chemical char- fate isolated from pig skin [9] and umbilical cord [8] acteristics of the ground substance and thereby alter has been shown to have a hybrid molecular structure the transport of molecules passing through it. In terms in which glucuronic acid is substituted for iduronic of steric effects, sulfated polymers would be stiffer than acid in the polymer chain, which renders these link- nonsulfated polymers and more extended in configura- ages susceptible to enzymatic cleavage by hyaluroni- tion, thereby increasing the physical domain which dase and other /J(l-»4)-glycosidases. Determination of they occupy and also the intramolecular spaces of the molecular weights before and after digestion with hy- gel. This would result in increased efficiency in the aluronidase therefore probes the polymer for its glucu- exclusion of large macromolecules from the gel, and ronic acid substitutions. hence in their more rapid transport by bulk flow be- Hyaluronidase did not alter the molecular size of tween tissue compartments. Macromolecules of inter- dermatan sulfate isolated from hog mucosa. In con- mediate size which can partly penetrate the gel would trast, treatment with hyaluronidase produced smaller behave differently. Their passage by bulk flow would molecular weight fragments of dermatan sulfate from be slowed but once in the gel they would filter through both the young and term human placenta. This indi- it more quickly. Transport of small molecules by these cated that both polymers had glucuronic substitutions. two mechanisms of bulk flow and molecular sieving The significantly greater reduction in molecular would not be altered by increased sulfation. On the weight of dermatan sulfate from the young placenta other hand, the increase in anionic charge would be compared with term placenta suggests three possibili- expected to increase the binding capacity of GAG for ties: (2) that the polymer from the young placenta cationically charged molecules, regardless of size, re- contains greater numbers of glucuronic acid substitu- tarding their passage through the matrix. The net ef- tions, (2) that the substitutions are positioned near the fects of these physical and chemical changes as the middle of the polymer chain, or (3) that the hybrid placenta ages might be summarized by decreased trans- molecules of each species have the same frequency and fer of small cationically charged molecules with rela- pattern of substitutions but there is a far greater pro- tively faster transport of most macromolecules by bulk portion of hybrids in the fetal material than exists in flow. the term. Some data has been generated in this laboratory Molecular weight determinations by optical mixing which indicates that dibasic amino acids bind to puri- spectroscopy does not permit differentiation of these fied chondroitin 6-sulfate in solution. In addition, re- three possibilities, but it may be possible to do this in cent reports indicate that polylysine and polyarginine principle by characterization of the molecular weight are tightly bound to sulfated GAG [10, 11]. distributions of the samples from line shape analysis These considerations suggest that the molecular [17]. Chemical analysis to resolve this question could structure and/or composition of placental GAG may not be carried out because of insufficient sample. Limi- in part regulate the transport of small and large mole- tation in the amount of sample is a frequent problem cules across the placenta, thereby affecting fetal in structural studies of developing systems. The tech- growth. We are currently testing this possibility by 976 LEE, JAMIESON, AND SGHAFER analyses of placenta obtained from term infants who 13. JAMIESON, A. M., LEE, T.-Y., AND SCHAFER, I. A.: Dermatan are small for gestational age. sulfate: Determination of structural changes during development by optical mixing spectroscopy. Biophys. J., 13: 154A (1973). Summary 14. JAMIESON, A. M., AND WALTON, A. G.: Studies of viscous drag in binary liquid mixtures by quasielastic light scattering. J. Chemical, enzymatic and structural analysis of GAG Chem. Physics, 58: 1054 (1973). isolated from the human placenta between 12-18 15. KAPLAN, D., AND MEYER, K.: Ageing of human cartilage. weeks of gestational age and at term (40 weeks) show Nature, 183: 1267 (1959). age-related differences in composition and structure of 16. KAPLAN, D., AND MEYER, K.: Mucopolysaccharides of aorta at these polymers. The young placenta contains increased various ages. Proc. Soc. Exp. Biol. Med., 105: 78 (1960). concentrations of chondroitin, whereas the term pla- 17. KOPPEL, D. E.: Analysis of macromolecular polydispersity in intensity correlation spectroscopy: The method of cumulants. centa shows less unsulfated GAG but higher concentra- J. Chem. Physics, 57: 4814 (1972). tions of dermatan sulfate and heparan sulfate. 18. LAGUNOFF, D., AND WARREN, G.: Determination of 2-deoxy-2- Changes in the composition are paralleled by differ- sulfoaminohexosc content of mucopolysaccharides. Arch. ences in the molecular structure of chondroitin sulfate Biochem. Biophys., 99: 396 (1962). and dermatan sulfate. The biologic implication of 19. LAURENT, T. C: The exclusion of macromolecules from poly- saccharide media. In: G. Quintarelli: The Chemical Physiol- these data may be related to postulated roles of GAG ogy of Mucopolysaccharides, Chapt. 11. (Little Brown & Co., in the regulation of placental transport. Boston, 1968). 20. LAURENT, T. C: Sterical interaction between polysaccharides References and Notes and other macromolecules. Structure and Function of Connec- tive and Skeletal Tissue, p. 252 (Butterworths, London, 1965). 1. BARNES, A. D.: Placental transfer. In: Intra-utcrine Develop- 21. LONCO, L. D.: Disorders of placental transfer. In: N. S. Assali: ment, Chapt. 5 (Lea & Febiger, Philadelphia, 1968). Pathophysiology of Gestation (Academic Press, New York, 2. BITTER, T., AND MUIR, H. M.: A modified uronic acid carba- 1972). zole reaction. Anal. Biochem., 4: 330 (1962). 22. LOVELL, D., TIGHE, J. R., AND CURRAN, R. C: The chemical 3. BOAS, N. F.: Method for the determination of hexosamines in determination of acid mucopolysaccharides and collagen of tissues. J. Biol. Chem., 204: 553 (1953). the normal human placenta. J. Obstet. Gynaec. Brit. Com- 4. BREEN, M., WEINSTEIN, H. G., JOHNSON, R. L., VEIS, A., AND monw., 74: 568 (1967). MARSHALL, R. T.: Acidic glycosaminoglycans in human skin 23. LOWRY, O. H., ROSEBROUCH, N. J., FARR, A. L., AND RANDALL, during fetal development and adult life. Biochim. Biophys. R. J.: Protein measurement with the Folin phenol reagent. Acta, 201: 54 (1970). J. Biol. Chem., 193: 265 (1951). 5. CALATRONI, A., AND DI FERRANTE, N.: The glycosaminoglycans 24. LUDOWIEG, J., AND BENMAMAN, J. D.: Colorimetric differentia- of human term placenta. Carbohyd. Res., 10: 535 (1969). tion of hexosamines. Anal. Biochem., 19: 80 (1967). 6. CIFONELLI, J. A., LUDOWIEG, J., AND DORFMAN, A.: Chemistry of 25. MATI-IEWS, M. B.: Structural factors in cation binding to /3-heparin (chondroitinsulfuric acid B). J. Biol. Chem., 2}}: anionic polysaccharides of connective tissue. Arch. Biochem. 541 (1958). Biophys., 104: 394 (1964). 7. DODCSON, K. S., AND PRICE, R. G.: A note on the determination 26. MEYER, K.: Biochemistry and biology of mucopolysaccharides. of the ester sulfate content o£ sulfated polysaccharides. Bio- Amer. J. Med. 47: 664 (1969). chem. J., 84:106 (1962). 27. MEYER, K., DAVIDSON, E. A., LINKER, A., AND HOFFMAN, P.: 8. FRANSSON, L. A.: Structure of dermatan sulfate. III. The The acid mucopolysaccharides of connective tissue. Biochim. hybrid structure of dermatan sulfate from umbilical cord. J. Biophys. Acta, 21: 506 (1956). Biol. Chem., 24}: 1504 (1968). 28. PELZER, H., AND STAIB, W.: Steriodkonjugat. I. Hochspannungs- 9. FRANSSON, L. A., AND RODEN, L.: Structure of dermatan sulfate. papier-elektrophorese von 17-ketosteroidkonjuaten. Clin. I. Degradation by testicular hyaluronidase. J. Biol. Chem., Chim. Acta, 2: 407 (1957). 242: 4161 (1967); II. Characterization of products obtained by 29. ROBERTSON, W. V. B.: Personal communication. hyaluronidase digestion of dermatan sulfate. J. Biol. Chem., 30. SAITO, H., YAMAGATA, T., AND SUZUKI, S.: Enzymatic methods 242: 4170 (1967). for the determination of small quantities of isomeric chon- 10. GELMAN, R. A., AND BLACKWELL, J.: Mucopolysaccharide-poly- droitin sulfates. J. Biol. Chem., 243: 1536 (1968). peptide interaction: Effect of the position of the sulfated 31. SCHUBERT, M., AND HAMERMAN, D.: The ground substance of group. Biochim. Biophys. Acta, 297: 452 (1973). connective tissue. A Primer on Connective Tissue Biochem- 11. GELMAN, R. A., RIPPON, W. B., AND BLACKWELL, J.: Interac- istry, Chapt. 3 (Lea & Febiger, Philadelphia, 1968). tions between chondroitin sulfate C and poly-L-lysine: Pre- 32. SCOTT, J. E.: Ion binding in solutions containing acid muco- liminary report. Biochem. Biophys. Res. Commun., 48: 708 polysaccharides. In: G. Quintarelli: The Chemical Physiology (1972). of Mucopolysaccharides, Chapt. 12 (Little Brown & Co., 12. HARUKI, F., AND KIRK, J. E.: A method for separation of Boston, 1968). chondroitin sulfate B from isomers A and C by paper elec- 33. SVEJCAR, J., AND ROBERTSON, W. V. B.: Microseparation and trophoresis with zinc sulfate or zinc acetate solution as elec- determination of acid glycosaminoglycans (mucopolysaccha- trolyte medium. Biochim. Biophys. Acta, 136: 391 (1967). rides). Anal. Biochem., 18: 333 (1966). Developmental changes in placental glycosaminoglycans 977

34. TAKEUCHI, K., AND SCHAFER, I. A.: Glycosaminoglycan compo- cens, and chondroitinase ABC was prepared from Proteus sition of the human term placenta. Paediat. Univ. Tokyo, 18: vulgaris by Seikagaku Kogyo Co., Ltd., Tokyo, Japan, and 127 (1970). distributed by Miles Laboratories, Inc., Kankakee, 111. 35. Pronase B, 45,000 proteolytic U/mg, broad spectrum protease 40. The standards were 6-, 4-, and nonsulfated galactose deriva- isolated from Streptomyces griseus, Calbiochem, Los Angeles, tives of 2-acetamido-2-deoxy-3-0-(j3-D-gluco-4-enepyranasylu- Calif. ronic acid), prepared by Seikagaku Kogyo Co., Ltd., Tokyo, 36. Diastase (hog pancreas), 1/50/5, Nutritional Biochemicals Japan, and distributed by Miles Laboratories, Inc., Kankakee, Corporation, Cleveland, Ohio. 111. 37. Bovine testicular hyaluronidase, 385 USP U/mg, Worthington 41. Beckman Instruments, Spinco Division, Palo Alto, Calif. Biochemical Corporation, lot no. 2592 HSE, Freehold, N. J. 42. Precision Instrument Co., Philadelphia, Penn. 38. Standard GAG was supplied by Dr. Martin B. Mathews, Uni- 43. Supported by National Institutes of Child Health and Human versity of Chicago, Chicago, Illinois, and consisted of hyalu- Development Grant nos. 03488 and 05996, National Institute ronic acid and chondroitin-6-sulfate from human umbilical of Dental Research Grant no. DE 02587, and the Cleveland cord, chondroitin-4-sulfate from the notochord of rock Diabetes Association. sturgeon, dermatan sulfate and heparin from hog mucosal 44. Requests for reprints should be addressed to: I. A. SCHAFER, tissues, keratan sulfate from bovine corneal tissue, keratan M.D., Department of Pediatrics, Cleveland Metropolitan Gen- sulfate 2 from human costal cartilage, and heparin sulfate eral Hospital, 3395 Scranton Rd., Cleveland, Ohio 44109 from bovine lung. (USA). 39. Chondroitinase AC II was prepared from Arthrobacter aures- 45. Accepted for publication August 1, 1973.